Landing an aerial vehicle is a very challenging problem. In operations involving aircraft, such as helicopters and airplanes, it is critical to detect landing zones and perform safe landing operations. However, landing operations can be difficult in numerous situations, such as emergency landings or under low visibility conditions. Pilots spend numerous hours practicing touchdowns because of the risk involved during landing phase. While autonomous landing technologies have been an active area of research, autonomous landing zone detection capabilities are currently limited to looking for large flat areas. As such, to an autonomous landing zone detection system, most of an airport would look like a feasible site. This means the system may choose to land on an inappropriate location, such as an active taxiway or a patch of grass, where air traffic controllers may not approve. Therefore, while current systems may be capable of detecting potential landing zones, the results may not be practical.
Conventional methods and systems have generally been considered satisfactory for their intended purpose. However, there is still a need in the art for systems and methods that provide for improved detection of landing zones. The present disclosure provides a solution for this need.